Dubnau 1999 ARM (Review - DNA Uptake in Bacteria)

8/3/2019 Dubnau 1999 ARM (Review - DNA Uptake in Bacteria)

1/29

?Annu. Rev. Microbiol. 1999. 53:21744

Copyright c 1999 by Annual Reviews. All rights reserved.

DNAU PTAKE IN BACTERIADavid DubnauPublic Health Research Institute, New York, NY 10016;e-mail: [email protected]

Key Words competence, transformation, DNA transport, type-2 secretion, pilusformation

s Abstract Natural competence is widespread among bacterial species. The mech-anism of DNA uptake in both gram-positive and gram-negative bacteria is reviewed.The transformation pathways are discussed, with attention to the fate of donor DNAas it is processed by the competent cell. The proteins involved in mediating varioussteps in these pathways are described, and models for the transformation mechanismsare presented. Uptake of DNA across the inner membrane is probably similar in gram-positive and gram-negative bacteria, and at least some of the required proteins areorthologs. The initial transformation steps differ, as expected, from the presence of an outer membrane only in the gram-negative organisms. The similarity of certainessential competence proteins to those required for the assembly of type-4 pili and fortype-2 protein secretion is discussed. Finally several hypotheses for the biological roleof transformation are presented and evaluated.

CONTENTS

Denitions and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218The Uptake Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218

Uptake of DNA in Gram-Positive Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218Uptake of DNA in Gram-Negative Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Competence Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222DNA Receptor Protein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222Competence Proteins Related to Those Involved in Type-4 Pilus

Assembly and Secretion: The PSTC Proteins . . . . . . . . . . . . . . . . . . . . . . . . 223Role of the PSTC Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227A Competence Nuclease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227Competence Proteins Required for Transport . . . . . . . . . . . . . . . . . . . . . . . . . 228Additional Competence Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229Energetics of DNA Uptake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230Models for DNA Uptake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

byCopenhagenUniversityon11/15/10.Forpersonaluseonly.


2/29

?218 DUBNAU

DEFINITIONS AND SCOPE

Natural competence is a genetically programmed physiological state permitting

the efcient uptake of macromolecular DNA. It is distinct from articial trans-formation involving electroporation, protoplasts, and heat shock/CaCl 2 treatment.This review deals only with natural competence. Recombination is not discussed.In many bacteria, competence is highly regulated, and much research has beendevoted to exploring the complex control mechanisms involved. This regulatorywork is not included here, and discussion is limited to the mechanism of DNAuptake and to the evolutionary signicance of competence.

THEUPTAKEPATHWAY

In gram-positive bacteria, DNA must pass through the cell wall and the cyto-plasmic membrane. In gram-negative bacteria, DNA must also traverse the outermembrane. Additional steps must therefore be involved in the gram-negative trans-formation systems, and the initial interaction of DNA with the cell surface mustbe different in the two types of bacteria. We begin with a description of the trans-formation pathway in thegram-positive organisms Bacillus subtilis and Streptococ-cus pneumoniae and then describe the gram-negative systems, with emphasis on Neisseria gonorrhoeae and Haemophilus influenzae . These initial descriptions arerestricted to the fate of donorDNA. Discussion of the proteins involved is presentedlater. A summary of the transformation pathways is provided in Figure 1. Becausethese pathways were described in the earlier literature and have been extensivelyreviewed (32, 61, 76, 88, 128), they are presented briey here.

Uptake of DNA in Gram-Positive Bacteria

Binding of DNA The rst step in transformation is the binding of double-

stranded DNA to the cell surface. In B. subtilis it has been estimated that there are50 binding sites per competent cell (36, 126). Binding occurs with no detectablelag (36,81) and without base sequence preference. Immediately after binding,intact double-stranded DNA can be recovered (38). In S. pneumoniae there alsoappears to be no base sequence specicity for transformation. It has been deter-mined that there are 3375 uptake sites per colony-forming unit in this organism(45). In both organisms the mass of DNA bound to the cell is proportional to themolecular size of the DNA (33, 69); binding occurs to a xed number of sites witha probability that is independent of molecular weight.

Fragmentation Within 30 sites of binding to B. subtilis , double-stranded DNA



3/29


4/29

?220 DUBNAU

strands are transported across the membrane, whereas the non-transported strandis degraded.

In B. subtilis , after a lag of 1 to 1.5 min at 37 C, bound donor chromosomal

DNA becomes inaccessible to added DNase (81). At thesame time, single-strandedDNA of donor origin can be recovered from lysed cells (28, 106). These appar-ently concomitant events indicate that the single-stranded DNA has been trans-ported into the cytosol. The average length of these single-stranded fragments is615 kb (28). This compares reasonably with the length of the double-strandedfragments recoverable from the cell surface (13.518 kb). These single strandshave not yet formed stable base-paired complexes with the recipient chromosome.Simultaneously with the appearance of internalized single-stranded DNA and theacquisition of DNase resistance, acid-soluble degradation products appear in the

medium (35). Theseproducts consist of 5 -nucleotides, nucleosides, and free bases.Because nucleotides cannot ordinarily pass freely across membranes, it is likelythat the nuclease responsible for this degradative step is localized outside the mem-brane or within an aqueous channel. The internalized strands interact efcientlywith complementary sequences in the recipient chromosome to yield heterodu-plex DNA. An unligated complex in which the donor and recipient moieties arestabilized only by base pairing has been detected (6, 37). The average size of the donor moiety in this transient complex was estimated as 8.3 kb. The averagesize of the nal integrated donor DNA was 8.510 kb (34, 43).

In B. subtilis the size correspondence between the surface-localized fragments,the internalized single strands, and the non-covalently bound and nal integrateddonor segments is satisfying. This correspondence is consistent with thehypothesisthat these various molecular forms are related as precursors and products andthat they are formed from one another in the order listed. Kinetic analysis hasestablished that this is the case for the surface localized fragments, the internalizedsingle strands, and the integrated donor fragments (29, 36). Without such evidenceone or more of these forms could be the products of side reactions.

As noted above, single genetic markers become DNase resistant after a 1- to

1.5-min lag during the transformation of B. subtilis . However, marker pairs requireadditional time, which increases with the physical distance between the markers(130, 131). This indicates that DNA is taken up in a linear fashion. Based onthis data and on newer sequence information, we estimate that DNA passes intothe cytosol at the rate of 180 nucleotides/s at 28 C, which was the tempe-rature used for the marker pair uptake experiment. Linear uptake ts nicely withthe existence of an aqueous channel for the transport of DNA across the mem-brane. The nature of the lag preceding transport of a single genetic marker isunclear, although it must include the time needed for both cleavage and for trans-

port across the membrane of the minimal segment adequate for recombination.This should take less than a minute at 28 C. However, the lag measured at this



5/29

?DNA UPTAKE IN BACTERIA 221

Transport of DNA in S. pneumoniae is similar to that in B. subtilis in mostrespects. Linked pairs require more time for entry than do single markers, againimplying linear uptake (49). Single strands appear in the cells (68, 94), and one

strand equivalent is released into the medium as low-molecular-weight products(74, 77, 98). These consist of short oligonucleotides (74). It is believed that trans-port initiates at a second break opposite the initial single-strand nick (70). Thepolarity of DNA entry in this organism is 3 5 (92). In an important study carriedout with radiolabeled linear and circular donor molecules, M ejean & Claverys(93) found that the degradation of the non-transported strand proceeded with 5 3polarity and at approximately the same rate as the entry of the transported strand.The uptake rate was estimated as 90100 nucleotides/s at 31 C. It is temptingto conclude from these results that the degradation and transport processes are

mechanistically coupled, although this is not certain.

Uptake of DNA in Gram-Negative Bacteria

The H. influenzae and N. gonorrhoeae uptake pathways are closely related. Ef-cient DNA uptake by these gram-negative bacteria requires the presence of aspecic uptake sequence (25, 27, 39, 50, 127). Uptake sequences are present atfrequencies far greater than those expected randomly. Uptake sequences are of-ten found as inverted repeats between genes, acting as transcriptional terminators(129). Inspection of the H. influenzae genome revealed an extended 29-base-pair(bp) consensus uptake sequence. However, not all gram-negative competence sys-tems display such specicity; Acinetobacter calcoaceticus is able to take up DNAfrom any source (87, 103). Another major difference between the gram-positiveand -negative systems is the presence of an outer membrane in the latter.

Binding A binding step, corresponding to the initial irreversible attachment of DNA to the cell surface in a form accessible to DNase or to shearing forces, hasnot been characterized directly in the H. influenzae and N. gonorrhoeae systems,

perhaps because of the rapidity with which DNase resistance is acquired. Anuptake rate of 5001000 nucleotides/s has been measured (61). In H. influenzaethe number of active sites for DNA uptake is only 48 per competent cell (30).

Fragmentation Circular 11.5-kb plasmid DNA was used to transform N. gonor-rhoeae (9). After the acquisition of DNase resistance, double-stranded moleculeswere recovered from the cells, which had been linearized at random positions.When linear plasmid DNA was used, no additional cleavage events were detected.However, when a larger plasmid DNA (42 kb) was used, more extensive double-

strand cleavage was observed. This pattern is reminiscent of that observed withthe gram-positive systems, and a similar mechanism may exist to introduce endsf b h i b



6/29

?222 DUBNAU

in the gram-negative systems. The acquisition of DNase resistance in these sys-tems is not concomitant with the conversion to single-stranded DNA. Instead,DNase resistance probably corresponds to entry into the periplasmic compartment

or into a specialized structure. In H. influenzae , a particle was observed on thesurface of competent cells (61). These transformasomes were considered to bestructures in which double-stranded donor DNA is sequestered early in the trans-formation pathway. This important work has not been pursued in recent years, andtransformasomes deserve re-evaluation.

Transport In H. influenzae , transformation results in single-strand integration(101), and it is likely that one strand is degraded during transport across the innermembrane (61) just as in the gram-positive systems. However free cytoplasmic

single-stranded DNA of donor origin has not been detected in this organism. In-stead it is believed that only a short length of single-stranded DNA is present at agiven time as the DNA crosses the inner membrane and searches out a complemen-tary sequence in the recipient. The degradation of the nontransported strand maybe concomitant with transport and integration. This would imply the existence of a degradation-transport-recombination protein complex associated with the innermembrane. Transport proceeds in a 3 5 direction (8), as it does in S. pneumo-niae . Recently a low level of single-stranded donor DNA was detected duringgonococcal transformation (16). It is not certain that this material is a precursor

of integrated DNA.

Acinetobacter Transformation The transformation of the gram-negative A. cal-coaceticus in some respects resembles that of the gram-positive systems. It wasconcluded that this organism takes up DNA from any source (103), and an initialbinding step resulting in DNase-sensitive association of DNA with the cell surfacewas identied (107). Moreover, in A. calcoaceticus double-stranded molecules areconverted to single strands on transport (103).

COMPETENCEPROTEINS

Severalgenes andproteins required forDNA uptakehavebeen characterized. Manyare conserved in the gram-positive and gram-negative systems. These proteinsand their individual roles are now described. After these descriptions, an attemptis made to integrate this information with our knowledge of the transformationpathways to construct summarizing models for gram-positive and gram-negativetransformation.

DNA RECEPTOR PROTEIN



7/29


open reading frame of the comE operon (54, 60). Non-polar null mutations incomEA eliminate DNA binding. An in-frame deletion in comEA results in theexpression of a shortened protein that is detectable in western blots (60). These

mutants are still able to bind DNA about half as well as the wild type, but they failto transport DNA. Thus ComEA has more than one function in transformation.ComEA possesses a single membrane-spanning domain near its N terminus andis an integral membrane protein with its C terminus outward (60). ComEA isintimately associated with the cell wall, presumably via its C-terminal domainbecause it can be chemically cross-linked to wall material in vivo (YS Chung &D Dubnau, unpublished data).

A His6-tag was substituted for the membrane-spanning domain of ComEA,and the resulting protein was puried (109). The His 6-ComEA protein bound to

double-stranded DNA in gel shift assays, with an apparent K d of 5 10 7

M.Membrane proteins from competent cultures were tested for DNA binding insouthwestern assays, and it was shown that intact ComEA was capable of bind-ing DNA. In both assays, a marked preference for double- over single-strandedDNA was noted, and binding occurred with no apparent sequence specicity. TheC terminus of ComEA shows similarity to other nucleic acid-binding proteinsand contains a possible helix-turn-helix motif (60, 109). A puried protein lackingthis motif had no detectable DNA-binding activity in the gel shift assay. Theseproperties of ComEA imply that it is a DNA receptor for transformation.

A ComEA ortholog has been detected in S. pneumoniae and shown to be essen-tial for transformation (15, 104). Proteins possessing similarity to the C-terminalDNA-binding domain of ComEA are widespread in nature, including those in H. influenzae and N. gonorrhoeae (109). However, there is no evidence that theseproteins are involved in transformation, and, because DNA uptake in these organ-isms requires a specic sequence, a ComEA ortholog may not be involved. It willbe interesting to see whether such a protein is needed by A. calcoaceticus in view of the similarity of its transformation pathway to that of the gram-positive organisms.

COMPETENCEPROTEINS RELATED TO THOSEINVOLVED IN TYPE-4 PILUS ASSEMBLYAND SECRETION: The PSTC Proteins

Another group of proteins encoded by the comG operon and by comC was shownto be required for DNA binding in B. subtilis (1, 53, 96). Analysis of non-polarmutants lacking each of the seven ComG proteins demonstrated that they areindividually needed for binding (19). In their absence, transformation was unde-

tectable. These proteins resemble a widespread group required in gram-negativebacteria for the assembly of type-4 pili, for the type-2 secretion pathway, and for



8/29


9/29


which is thought to require pilus retraction (12). The reverse is not true; PilT of N. gonorrhoeae (143) and Pseudomonas aeruginosa (141) is needed for twitchingmotility but not for pilus formation. PSTC proteins fall into ve classes, which

are discussed in turn below. The ubiquity of PSTC genes and their use for DNAand protein transport and for pilus assembly in diverse species suggest an an-cient mechanism, which predates the divergence of gram-positive and -negativebacteria.

Class 1, exemplied by ComGA in B. subtilis (1), consists of membrane-associated proteins with consensusnucleotide-binding sites.Members of this grouphave been implicated in transformation of both gram-negative (52, 107, 143) andgram-positive bacteria (15, 90, 104) as well as in pilus assembly (102), twitchingmotility (141, 143), and type-2 protein secretion (108). Class-1 proteins are also

required for conjugation-related systems. One of these, TrwD, required for conju-gation of the plasmid R388, has been shown to be an ATPase (116). The B. subtilisclass-1 PSTC protein (ComGA) is located as a peripheral membrane protein onthe cytosolic face of the membrane (18).

ComGB of B. subtilis exemplies a second class, consisting of membraneproteins apparently with three predicted membrane-spanning segments. Class-2proteinshavebeen implicated in pilusassembly (102)and type-2 secretion (108), aswell as in competence in gram-positive (1, 15, 90, 104) and gram-negative systems(52,138).

The third class consists of small proteins with conserved sequences at theirN-termini (24, 15, 104, 107, 115, 123, 124). Representatives of this class are re-quired for competence, secretion, and pilus assembly. Four of the B. subtilis ComGproteins (ComGC, ComGD, ComGE, and ComGG) and two proteins from thestreptococcal systems possess such conserved sequences (15,90, 104), resemblingthe cleavage sites of type-4 prepilin proteins. The major pilin protein of N. gonor-rhoeae is a member of this class and is required for transformation (124), as is an Acinetobacter protein (107). In nearly all of the systems that have been studied,multiple class-3 proteins are needed. An exception may be H. influenzae , in which

a single class-3 protein was detected in a search of the genome (D Dubnau, un-published data). Only for type-4 pilus assembly are the functions of any of theseproteins understood; one of them is the precursor of the major structural proteinof the pilus. In general the conservation of sequence among the members of thisclass is restricted to the hydrophobic N-terminal domain. However members of one subgroup required for competence, consisting of the B. subtilis ComGC pro-tein, ComYC from S. gordonii (90), and CglC from S. pneumoniae (15,104), aresimilar over their entire lengths.

The four B. subtilis class-3 pre-proteins (ComGC, ComGD, ComGE, and

ComGG) are processed by the peptidase, ComC (18, 20). They are integral mem-brane proteins, anchored by their single predicted N-terminal membrane-spanning



10/29

?226 DUBNAU

ComGC, ComGD, ComGG, and possibly ComGE, undergo translocation and areno longer located as integral membrane proteins (18, 20). This requires the pepti-dase (ComC) and is probably dependent on processing. Consequently about one

fourth of each mature class-3 protein is peripherally associated with the outer faceof the membrane, and another fourth of the total is released from the protoplaston removal of the cell wall. We have recently found that these proteins can beefciently cross-linked in vivo to cell wall material (YS Chung & D Dubnau, un-published data). Additional cross-linking experiments have shown that portions of both the mature and unprocessed ComGC are probably present as homodimers.The single cysteine residue of ComGC is apparently involved in vivo in an in-tramolecular disulde bond. A portion of ComGG, on the other hand, is in ahomodimer that is stabilized by a disulde bond (18).

The fourth class of proteins in this group consists of membrane-localizedpeptidase/transmethylases. These cleave the class-3 proteins at a site within theN-terminal conserved sequence and, in at least some cases, transfer a methylgroup from S-adenosyl methionine to the newly formed primary amino group atthe N-terminus (133). Processing of the class-3 B. subtilis competence proteins isdependent on the peptidase ComC (18, 20).

The class-5 PSTC proteins are the secretins, which exhibit sequence similar-ities in their C-terminal domains (121). Secretins are required for competence(31, 136), pilus assembly (31, 91), and type-2 secretion (24). Secretins are also

involved in lamentous phage maturation (122) and in the so-called type-3 secre-tion systems (62). These proteins are located in the outer membrane, where theyform large multisubunit complexes (17, 56, 65, 100, 125). Protein pIV, the best-studied secretin, is required for the maturation of the lamentous phages. pIVforms a cylindrical complex, composed of 14 identical subunits, with a centralpore (82). Other secretins form similar structures (10, 23, 66). Secretins are almostcertainly involved in permitting passage of phage particles, DNA, protein, or piliacross the outer membrane. It is not surprising that they have been implicated ingram-negative (31, 136) but not in gram-positive transformation systems. A search

of the completed B. subtilis genome failed to indicate the presence of a secretin(D Dubnau, unpublished data).

An additional protein, PilC, is required forcompetence in N. gonorrhoeae (119).Although PilC is pilus-associated, it is not absolutely required for pilus production(120) and does not resemble any of the PSTC proteins described above. Interest-ingly, competence can be partially restored to a pilC null mutant by the additionof puried PilC protein or, even better, by the addition of crude preparations of pili-containing PilC (119). These results strongly suggest that PilC functions onthe cell surface. In A calcoaceticus , a PilC ortholog has also been identied as

a competence protein (83). In both organisms this protein is required for DNAuptake to a DNase-resistant state (83, 119). In its absence, DNA binding to the



11/29


12/29


13/29


(53, 84). In contrast, knockouts of other B. subtilis competence genes decreasetransformation at least 10 7-fold. In the absence of ComFA, the other competenceproteins may be able to position the end of an incoming strand near the aqueous

channel. Entry might then occur inefciently via diffusion, perhaps directionallybiased by interaction with single-strand binding protein (SSB) in the cytosol. AnSSB ortholog (CilA) is required for transformation in S. pneumoniae (15). A comF -like operon preceded by a competence regulatory signal [a cin box (15)] was alsodetected in S. pneumoniae (21). A PriA ortholog was identied in H. influenzae(42), and proteins with similarity to ComFA were also detected in a search of the incomplete N. gonorrhoeae and N. meningitidis databases (D Dubnau, unpub-lished data). No data exist concerning the possible roles of any of these proteinsin transformation, and it is uncertain whether a ComFA equivalent is an essential

competence protein in the gram-negative systems.The comFA and comEC null mutations and the in-frame deletion of comEA ,which prevents transport without a major effect on binding, also prevent therelease of acid-soluble products from radiolabeled donor DNA (R Provvedi &D Dubnau, unpublished data). These results suggest that transport may be neces-sary for degradation, although the nuclease is most likely located outside the mem-brane, consistent with the appearance of the acid soluble products in the medium(35). Transport may serve to drag the incoming DNA past the nuclease activesite.

ADDITIONALCOMPETENCEPROTEINS

Por, a periplasmic protein disulde oxidoreductase, is required for the transfor-mation of H. influenzae (137). In its absence, DNA binding does not occur. Por islocated in the periplasm and is required for competence-associated changes in theprotein composition of the membrane. It is likely that Por is needed for the correctfolding of one or more competence proteins. DprA is another protein required for

transformation in H. influenzae (63, 64), as well as in S. pneumoniae (15) and B.subtilis (P Tortosa & D Dubnau, unpublished data). In Haemophilus , the dprAmutant took up DNA into a DNase-resistant form as readily as the wild type, buttransformation was severely depressed. DprA may be needed to transport DNAacross the inner membrane.

As noted above, SSB (CilA) has been shown to be a competence protein inS. pneumoniae (15). A deletion mutant lacking the 33 C-terminal amino acidresidues exhibited 10% of the wild-type transformability. It is possible that SSBplays a supporting role in uptake. For instance, as the 3 donor DNA terminus

enters the cytosol, SSB may bind, facilitating transport and formation of a DNA-RecA lament before integration. In the absence of SSB, transport and lament



14/29

?230 DUBNAU

Inactivation of ComFC in B. subtilis decreases transformation only 5- to 10-fold(84), does not affect DNA binding, and decreases transport only slightly (85).This protein may affect a step that follows transport. In H. influenzae , a knockout

mutant of com-101 bound DNA as well as the wild type, but it released much lessdonor DNA in acid-soluble form (78). It is difcult to reconcile this phenotypewith the one observed in B. subtilis .

Two competence genes in N. gonorrhoeae may be involved in murein metab-olism. comL encodes a protein that is associated with and perhaps covalentlyattached to peptidoglycan (46). Complete loss of comL is lethal, but a pleiotropicmutation was characterized that decreased both competence and cell volume. Itappears that the N-terminal half of ComL, a lipoprotein, is sufcient to support via-bility but not competence. No ComL ortholog is encoded by B. subtilis (D Dubnau,

unpublished data), although a similar protein (HI0177) is encoded by H. influenzae(42). Inactivation of tpc , another N. gonorrhoeae gene, is also pleiotropic, resultingin transformation deciency and a cell separation defect (47). Extracts of a tpc mu-tant exhibited slightly decreased murein hydrolase activity. No obvious orthologof Tpc exists in the database. It was suggested that ComL and Tpc might be mureinhydrolases, responsible for providing access for DNA through the wall (46, 47).

ENERGETICS OF DNA UPTAKE

In B. subtilis , transformation is inhibited by a number of uncoupling reagents,and the extent of inhibition paralleled the effect of these reagents on membranepotential (51). Inhibition occurred at or before the transport step because the ac-quisition of DNase resistance was inhibited. The addition of arsenate to the trans-formation medium halved the ATP pool but had no effect on transformation. It wasconcluded that the ... initial stages of genetic transformation proceeded despitethe drastically lowered value of the intracellular phosphorylation potential (51).However, this conclusion must be questioned, because measurement of intracellu-

lar ATP was carried out with the bulk culture, whereas only 10% of the cells ina B. subtilis culture become competent for transformation. It was also concludedthat both the pH gradient and the membrane potential are required for DNA up-take (51). Grinius therefore proposed that DNA binds to the cell surface, forminga nucleoprotein complex that binds protons and acquires a positive charge. Thecomplex then electrophoreses across the membrane, releasing the protons. How-ever, another study with inhibitors (140) has concluded that pH alone is requiredfor DNA transport in B. subtilis , suggesting a proton symport mechanism for DNAuptake. In S. pneumoniae , it was proposed that degradation of one strand provides

the driving force for introduction of the intact strand (118). However, this modeldoes not readily accommodate the observation that single-strand transformation



15/29


state (D Dubnau, unpublished data). In H. influenzae , it was concluded that bothcomponents of the proton motive force could drive DNA uptake (14).

Although transformation requires the input of metabolic energy, no clear picture

exists as to the proximal source of this energy. Because all the relevant experimentsso far involve the action of inhibitors on whole cells, the available information isindirect and subject to various interpretations. For instance, is proton motive forcerequired to drive DNA transport directly or to maintain the active state of anessential competence protein?

MODELS FOR DNA UPTAKE

The information presented above can besummarized in theform ofworking modelsthat ascribe particular roles to some of the competence proteins. Before presentingthese models, it is useful to compare transformation to the type-2 secretion systemsin gram-negative bacteria (110). In type-2 secretion, an unfolded substrate pro-tein is transported by the Sec system across the inner membrane. These substrateprotein molecules are delivered to the periplasm, where they are folded and thentransported across the outer membrane. During transformation of gram-negativebacteria, double-stranded DNA must be transported across the outer membranein the reverse direction. Conversion to single-stranded DNA and transport acrossthe inner membrane ensues by using proteins that are functionally analogous tothe Sec proteins. Transport across the inner membrane, probably through water-lled channels, requires that the substrate macromolecules be in particular exiblestatesunfolded for proteins and single-stranded for DNA. In the transformationand secretion systems, PSTC proteins appear to be required to provide accessthrough the wall and outer membrane and therefore to fulll similar roles in all of these systems.

The Gram-Positive Model

Figure 2 presents a working model for the transformation process in gram-positivebacteria. It draws on evidence taken from both S. pneumoniae and B. subtilis , al-though most of the information concerning the roles of individual proteins relieson the latter. Double-stranded DNA binds to the C-terminal domain of membrane-anchored ComEA. A nuclease then cleaves the bound DNA near the point of contact with ComEA. The newly formed DNA terminus is then delivered to themembrane-associated transport apparatus, containing ComEC and ComFA. A nu-clease that degrades the non-transported strand (EndA in S. pneumoniae ) is pre-

sumably also part of this apparatus. ComEA also plays a role in transport, possiblyby contributing structurally to the transport apparatus or perhaps by playing an



16/29

?232 DUBNAU

Figure 2 Transformation in gram-positive bacteria. The protein nomenclature is from Bacillus subtilis , but it is likely that a similar mechanism operates in Streptococcus

pneumoniae . Double-stranded DNA binds to the receptor protein ComEA. A double-strand cleavage event is followed by delivery to a nuclease (N) that degrades one strand,releasing acid-soluble products into the medium. The nuclease has not been identied in B. subtilis , whereas in S. pneumoniae it is encoded by endA . The ComG proteins are de-picted, forming a structure that provides access of DNA to ComEA through the wall. Onepossible model for the delivery step is shown in which ComEA conformation is alteredso that the bound DNA can contact the nuclease. The single-strand product of nucleaseaction then contacts the transporter protein ComFA, and the energy of ATP hydrolysis isused to drive the DNA into the cell. We postulate that ComEC forms an aqueous pore in

the membrane for this transport step. References and further details are provided in thetext.

that traverses the wall, permitting access of DNA to the ComEA receptor. It ispossible that these proteins serve instead to increase the porosity of the wall or thatthe individual ComG proteins play different roles, some increasing porosity andsome composing a wall-traversing structure.

The delivery process deserves comment. One possibility is that the DNA-binding domain of ComEA is close to the transport machinery. Another is that

some movement is required to establish this proximity. Two suggestive featuresare consistent with this idea. First, ComEA is similar to the C-terminal domain



17/29


membrane (18, 85), contain nucleotide-binding sites. The binding site of ComFAis known to be required for transport (86). Perhaps one of these proteins drives aconformational change in ComEA that delivers DNA to the transport pore. It is

interesting that just upstream from the DNA-binding domain of ComEA is a ex-ible sequence (QQGGGG). A similar sequence is conserved in the S. pneumoniaeortholog. When this was deleted, the mutant cells were able to bind but not totransport DNA (60). Perhaps bending of ComEA delivers DNA to the uptake ap-paratus. A second suggestive phenomenon is that of twitching motility, associatedwith type-4 pili. It is believed that this form of bacterial motion requires the as-sembly and disassembly of pili, and an ortholog of ComGA has been implicatedin the disassembly process (141). It is possible that the proposed wall-traversingstructure disassembles by an analogous process, and that this permits ComEA to

bend and thereby deliver DNA to the transport machine. Because disassembly of this structure and delivery of the bound DNA to the uptake machinery would betime dependent, this type of mechanism may explain the lag observed before theacquisition of DNase resistance in B. subtilis .

In all of the well-studied transformation systems, a ComEC ortholog is requiredfor DNA transport. These are polytopic membrane proteins and may form uptakepores (Figures 2, 3). In B. subtilis and probably in S. pneumoniae as well (21),ComFA is a component of this machinery, perhaps mobilizing the energy of ATPhydrolysis for this purpose. As DNA enters the cytosol it probably associates with

SSB and RecA, and these binding energies may facilitate uptake.

The Gram-Negative Model

Figure 3 presents a model for transformation in H. influenzae and N. gonorrhoeae .As noted above, transformation in A. calcoaceticus may be different, resemblingthat of the gram-positive bacteria. In N. gonorrhoeae the phenotypes associatedwith loss-of-function mutations in several competence genes have been analyzed(40, 48). PilC and the secretin protein PilQ are needed for outer membrane trans-

port, the latter to form a pore. The proposed murein hydrolases Tpc and ComLmay aid access through the wall. By analogy with the B. subtilis results describedabove, the PSTC proteins allow DNA passage across the wall and periplasm. Be-cause these proteins are needed for binding in N. gonorrhoeae (40), access to aDNA receptor analogous to ComEA may occur within the proposed PSTC proteincomplex. Double-strand cleavage by an unknown nuclease also occurs (9). ComA[in N. gonorrhoeae (41)] or Rec2 [in H. influenzae (22)] orthologs of the B. sub-tilis ComEC protein mediate transport across the inner membrane. Conversion tosingle-stranded DNA and degradation of one strand equivalent areprobably closely

coupled in time, if not mechanistically, to inner membrane transport. Because largeamounts of free single-stranded DNA are not detected, the integration step may



18/29

?234 DUBNAU

Figure 3 Transformation in gram-negative bacteria. The protein nomenclature is takenfrom the Neisseria gonorrhoeae system, but the mechanism is likely to be similar in Haemophilus influenzae . Double-stranded DNA enters the periplasm through an outermembrane (OM) pore consisting of the secretin protein PilQ. PilC, a pilus-associatedprotein, is required for this step. It is believed that the pilin protein (PilE) forms a structurethat traverses themurein layer and that remodeling of thewall for transformation requiresthe proteins ComL and Tpc. The DNA receptor proteins and the nuclease that degrades

one strand have not been identied. The intact single strand is then transported across theinner membrane (IM) through an aqueous pore, possibly formed by ComA, a ComECortholog. The transporter protein (T) has not been identied. References and furtherdetails are provided in the text.

WHAT USEIS COMPETENCE?

Competence is widespread. A review published ve years ago listed > 40 naturally

transformable bacterial species (88). Transformation requires more than a dozenproteins and is often exquisitely regulated, as if the expression of this capabilitymust be nely tuned to the needs of each organism. What is the selective forcethat has shaped and maintained these elaborate systems in so many species? Sev-eral hypotheses have been advanced, and as is usual with evolutionary argumentsthey are not easily testable and may not be mutually exclusive. These hypothe-ses can be characterized as DNA as food, DNA for repair, and DNA for geneticdiversity.

It was proposed that competence evolved to permit the uptake of DNA as a food

supply (113, 114), but this reviewer believes that this is not a major factor amongthe selective pressures that maintain the competence mechanism. For instance



19/29


Why would theelaborate transformation machinery evolve to meet thisneed, whichis met by a simpler and more generally useful pathway? It should be pointed outthat, because one strand equivalent is released into the medium, the competence

machinery discards half of the potential foodstuff, a wasteful mechanism. The H. influenzae and N. gonorrhoeae systems exhibit uptake specicity. This doesnot suggest a food-gathering mechanism. It cannot be ruled out that competencehas served such a function under some conditions, in some organisms, and perhapsat some point in evolutionary history, but this does not seem plausible as a generalselective force.

A second proposal is that transformation serves a function in DNA repair(59, 95, 142). Lysed cells provide DNA that is taken up and used for the repairof otherwise lethal lesions. This suggestion is supported by the nding that the

DNA repair machinery of B. subtilis is induced as part of the competence reg-ulon (55, 89). The repair hypothesis has been criticized because DNA-damagingagents do not induce competence (112). However, such induction is not a strongprediction of the repair hypothesis. Induction of a DNA repair system before theappearance of DNA damage could provide a selective advantage. Such an induc-tion mechanism may respond to conditions in which DNA damage is likely tooccur, not to the damage itself. For instance, as B. subtilis approaches stationaryphase, when competence is induced, its metabolism becomes more aerobic, andoxidative damage to DNA may be more likely. Such preventive induction might

repair damaged DNA before serious harm is done by the secondary introductionof double-strand breaks or the synthesis of toxic proteins. In B. subtilis , transfor-mation was found to increase survival in populations UV-irradiated before, but notafter, the addition of DNA (59, 95, 142). This was interpreted as favoring the DNArepair hypothesis. Similar data were reported for H. influenzae , but transformationwith a cloned fragment had the same effect on survival as total chromosomal DNA,indicating that repair was not targeted to the site of integration as predicted by therepair hypothesis (97). Foreign DNA had no effect. Somehow a recombinationevent increased the survival rate, raising important doubts about the validity of the

earlier experiments with B. subtilis .The third popular hypothesis proposes that transformation is a mechanism for

exploring the tness landscape. All genetic diversity ultimately derives from muta-tion, but recombination can generate new allelic combinations. Many examples of horizontal gene transfer exist, and it is likely that transformation has played a role.Two instructive examples have been reported recently. In N. meningitidis , sequenc-ing of the sodC gene revealed the presence of two H. influenzae transformationuptake sequences, which form the transcriptional terminator of this virulence de-terminant (67). The sequence of sodC resembled that of the Haemophilus ortholog

more closely than the sod gene of Escherichia coli , whereas other Neisseria geneswere more similar to those of E. coli than of H. influenzae . This was interpreted as



20/29

?236 DUBNAU

corresponding Haemophilus sequences contained a Neisseria uptake sequence,providing a plausible mechanism for the transfer of the associated DNA from Haemophilus to N. meningitidis .

Another example of evolution by transformation was discovered in Helicobac-ter pylori , a cause of human gastritis (134). The sequences of three DNA fragmentsfrom clinical strains of this organism were determined, and it was demonstratedthat polymorphisms in these loci are at linkage equilibrium. Because transforma-tion is the only known means of genetic exchange in this organism, it is likely thatcompetence is responsible for its panmictic population structure. It was pointed out(134) that recombination can contribute to genetic diversity in two ways, perhapsillustrated by the two examples given here. In H pylori , recombination may haveserved to prevent a reduction in diversity caused by selective sweeps (whereas

selection for a mutation that increases tness causes the loss of competing geno-types with their associated diversity) and founder effects associated with the initialspread of a small number of organisms to a new niche. Frequent recombinationwill rapidly disperse an advantageous mutation to many genetic backgrounds, pre-venting these bottleneck effects from reducing diversity. In the Neisseria example,on the other hand, transformation has apparently served to introduce new genesfrom another species.

Because transformation has actually enabled genetic exchange in natural pop-ulations, it is tempting to conclude that this explains the selection for competence.

But this remains speculative, and the debate concerning the evolutionary roleof transformation will continue. In fact the signals that induce competence arespecies-specic. This may be telling us that competence can serve a variety of needs and that each species has learned to use the ability to transport DNA to meetits particular requirements.

ACKNOWLEDGMENTS

I thank all the members of our lab forhelpfulcomments on the manuscript and espe-

cially for countless stimulating discussions. NIH Grants GM57720 and GM43756supported the work from our group.

Visit the Annual Reviews home page at http://www.AnnualReviews.org

LITERATURECITED

1. Albano M, Breitling R, Dubnau D. 1989.Nucleotide sequence and genetic organiza-

tion of the Bacillus subtilis comG operon. J. Bacteriol. 171:53864042 Alm RA Hallinan JP Watson AA Mattick

crease the similarity of type 4 mbriae tothe GSP protein secretion systems and pilY1

encodes a gonococcal PilC homolog. Mol. Microbiol. 22:161733 Alm RA Mattick JS 1995 Identication of



21/29


product possesses a pre-pilin-like se-quence. Mol. Microbiol. 16:48596

4. Alm RA, Mattick JS. 1996. Identicationof two genes with prepilin-like leader se-quences involved in type 4 mbrial biogen-esis in Pseudomonas aeruginosa . J. Bacte-riol. 178:380917

5. Archibald AR, Hancock IC, Harwood CR.1993. Cell wall structure, synthesis andturnover. In Bacillus subtilis and Other Gram-Positive Bacteria: Physiology, Bio-chemistry, and Molecular Biology , ed. ALSonenshein, R Losick, JA Hoch, pp. 381

410. Washington DC: Am. Soc. Microbiol.6. Arwert F, Venema G. 1973. Evidence for anon-covalently bonded intermediate in re-combining during transformation in Bacil-lus subtilis . In Bacterial Transformation ,ed.LJArcher, pp. 20314.New York: Aca-demic

7. Arwert F, Venema G. 1973. Transforma-tion in Bacillus subtilis . Fate of newly in-troduced transforming DNA. Mol. Gen.

Genet. 123:185988. Barany F, Kahn ME, Smith HO. 1983.

Directional transport and integration of donor DNA in Haemophilus influenzaetransformation. Proc. Natl. Acad. Sci. USA80:727478

9. Biswas GD, Burnstein KL, Sparling PF.1986. Linearization of donor DNA duringplasmid transformation in Neisseria gon-orrhoeae . J. Bacteriol. 168:75661

10. Bitter W, Koster M, Latijnhouwers M, deCock H, Tommassen J. 1998. Formation of oligomeric rings by XcpQ and PilQ, whichare involved in protein transport across theouter membrane of Pseudomonas aerugi-nosa . Mol. Microbiol. 27:20919

11. Bodmer W, Ganesan AT. 1964. Biochem-ical and genetic studies of integration andrecombination in Bacillus subtilis transfor-mation. Genetics 50:71738

12. Bradley DE. 1980. A function of Pseu-domonas aeruginosa PAO polar pili:

13. Breitling R, Dubnau D. 1990. A pilin-likemembrane protein is essential for DNAbinding by competent Bacillus subtilis . J. Bacteriol. 172:1499508

14. Bremer W, Kooistra J, Hellingwerf KJ,Konings WN. 1984. Role of the electro-chemical proton gradient in genetic trans-formation of Haemophilus inf luenzae . J. Bacteriol. 157:86873

15. Campbell EA, Choi S, Masure HR. 1998.A competence regulon in Streptococcus pneumoniae revealed by genomic analysis. Mol. Microbiol. 27:92939

16. Chaussee MS, Hill SA. 1998. Formation of single-stranded DNA during DNA trans-formation of Neisseria gonorrhoeae . J. Bacteriol. 180:511722

17. Chen LY, Chen DY, Miaw J, Hu NT. 1996.XpsD, an outer membrane protein requiredforprotein secretionby Xanthomonas campestris pv. campestris , forms a multimer. J. Biol. Chem. 271:27038

18. Chung YS, Breidt F, Dubnau D. 1998. Cell

surface localization and processing of theComG proteins, required for DNA bindingduring transformation of Bacillus subtilis. Mol. Microbiol. 29:90513

19. Chung YS, Dubnau D. 1998. All sevencomG open reading frames are required forDNA binding during the transformation of competent Bacillus subtilis . J. Bacteriol.180:4145

20. Chung YS, Dubnau D. 1994. ComC is re-quired for the processing and translocationof ComGC, a pilin-like competence pro-tein of Bacillus subtilis . Mol. Microbiol.15:54351

21. Claverys JP, Martin B. 1998. Competenceregulons, genomics and streptococci. Mol. Microbiol. 29:112627

22. Clifton SW, McCarthy D, Roe BA. 1994.Sequence of the rec-2 locus of Hemophilusinfluenzae : homologies to comE -ORF3 of Bacillus subtilis and msbA of Escherichiacoli . Gene 146:95100



22/29

?238 DUBNAU

requires the InvH lipoprotein for outermembrane localization. Mol. Microbiol.30:4756

24. dEnfert C, Reyss I, Wandersman C, Pugs-ley AP. 1989. Protein secretion by gramnegative bacteria. Characterization of twomembrane proteins required for pullu-lanase secretion by Escherichia coli K-12. J. Biol. Chem. 264:1746268

25. Danner DB, Deich RA, Sisco KL, SmithHO. 1980. An eleven-base-pair sequencedetermines the specicity of DNA up-take in Haemophilus transformation. Gene

11:3111826. Deleted in proof 27. Danner DB, Smith HO, Narang SA. 1982.

Construction of DNA recognition sites ac-tive in Haemophilus transformation. Proc. Natl. Acad. Sci. USA 79:239397

28. Davidoff-Abelson R, Dubnau D. 1973.Conditions affecting the isolation fromtransformed cells of Bacillus subtilis of high molecular weight single-stranded de-

oxyribonucleic acidofdonororigin. J. Bac-teriol. 116:14653

29. Davidoff-Abelson R, Dubnau D. 1973. Ki-netic analysis of the products of donordeoxyribonucleate in transformed cells of Bacillus subtilis . J. Bacteriol. 116:15462

30. Deich RA, Smith HO. 1980. Mecha-nism of homospecic DNA uptake in Haemophilus influenzae transformation. Mol. Gen. Genet. 177:36974

31. Drake SL, Koomey M. 1995. The productof the pilQ gene is essential for the bio-genesis of type IV pili in Neisseria gonor-rhoeae . Mol. Microbiol. 18:97586

32. DubnauD. 1993.Genetic exchangeandho-mologous recombination. In Bacillus sub-tilis and Other Gram-Positive Bacteria: Biochemistry, Physiology, and Molecular Genetics , ed. AL Sonenshein, JA Hoch,R Losick, pp. 55584. Washington, DC.:Am. Soc. Microbiol.

sis on the recombination mechanism. In Microbiology , ed. D Schlessinger, pp. 1427. Washington, DC: Am. Soc. Microbiol.

34. Dubnau D, Cirigliano C. 1972. Fate of transforming deoxyribonucleic acid afteruptake by competent Bacillus subtilis : sizeand distribution of the integrated donor se-quences. J. Bacteriol. 111:48894

35. Dubnau D, Cirigliano C. 1972. Fate of transforming DNA following uptake bycompetent Bacillus subtilis . III. Formationand properties of products isolated fromtransformed cells which are derived en-

tirely from donor DNA. J. Mol. Biol. 64:92936. Dubnau D, Cirigliano C. 1972. Fate of

transforming DNA following uptake bycompetent Bacillus subtilis . IV. The end-wise attachment and uptake of transform-ing DNA. J. Mol. Biol. 64:3146

37. Dubnau D, Cirigliano C. 1973. Fate of transforming DNA following uptake bycompetent Bacillus subtilis . VI. Non-

covalent association of donor and recipientDNA. Mol. Gen. Genet. 120:1016

38. Dubnau D, Cirigliano C. 1974. Uptake andintegration of transforming DNA in Bacil-lus subtilis . In Mechanisms in Recombi-nation , ed. RF Grell. London/New York:Plenum

39. Elkins C, Thomas CE, Seifert HS, Spar-ling PF. 1991. Species-specic uptake of DNA by gonococci is mediated by a10-base-pair sequence. J. Bacteriol. 173:391113

40. Facius D, Fussenegger M, MeyerTF. 1996.Sequential action of factors involved innatural competence for transformation of Neisseria gonorrhoeae . FEMS Microbiol. Lett. 137:15964

41. Facius D, Meyer TF. 1993. A novel deter-minant (comA ) essential for natural trans-formation competence in Neisseria gonor-rhoeae . Mol. Microbiol. 10:699712

42. Fleischmann RD, Adams MD, White O,



23/29


assembly of Haemophilus influenzae Rd.Science 269:496512

43. Fornilli SL, Fox MS. 1977. Electron mi-croscope visualization of the products of Bacillus subtilis transformation. J. Mol. Biol. 113:18191

44. Fox MS, Allen MK. 1964. On the mech-anism of deoxyribonucleate incorporationin pneumococcal transformation. Proc. Natl. Acad. Sci. USA 52:41219

45. Fox MS, Hotchkiss RD. 1957. Initiationof bacterial transformation. Nature 179:132225

46. Fussenegger M, Facius D, Meier J, MeyerTF. 1996. A novel peptidoglycan-linkedlipoprotein (ComL) that functions in nat-ural transformation competence of Neis-seria gonorrhoeae . Mol. Microbiol. 19:1095105

47. Fussenegger M, Kahrs AF, Facius D,Meyer TF. 1996. Tetrapac (tpc), a novelgenotype of Neisseria gonorrhoeae affect-ing epithelial cell invasion, natural trans-

formation competence and cell separation. Mol. Microbiol. 19:135772

48. Fussenegger M, Rudel T, Barten R, Ryll R,Meyer TF. 1997. Transformation compe-tence and type-4 pilus biogenesis in Neis-seria gonorrhoeae a review. Gene 192:12534

49. Gabor M, Hotchkiss R. 1966. Manifesta-tion of linear organization in molecules of pneumococcal transforming DNA. Proc. Natl. Acad. Sci. USA 56:144148

50. Goodman SD, Scocca JJ. 1988. Identi-cation and arrangement of the DNA se-quence recognized in specic transforma-tion of Neisseria gonorrhoeae . Proc. Natl. Acad. Sci. USA 85:698286

51. Grinius L. 1982. Energetics of gene trans-fer into bacteria. Sov. Sci. Rev. Sect. D 3:11565

52. Gwinn ML, Ramanathan R, Smith HO,TombJF. 1998. A newtransformation-de-cient mutant of Haemophilus inf luenzae

53. Hahn J, Albano M, Dubnau D. 1987. Iso-lation and characterization of Tn 917lac -generated competence mutants of Bacillussubtilis . J. Bacteriol. 169:31049

54. Hahn J, Inamine G, Kozlov Y, DubnauD. 1993. Characterization of comE , a latecompetence operon of Bacillus subtilis re-quired for the binding and uptake of trans-forming DNA. Mol. Microbiol. 10:99111

55. HaijemaBJ,van Sinderen D, WinterlingK,Kooistra J, Venema G, et al. 1996. Regu-lated expressionof thedinR andrecAgenesduring competence development and SOS

induction in Bacillus subtilis . Mol. Micro-biol. 22:758556. Hardie KR, Lory S, Pugsley AP. 1996. In-

sertion of an outer membrane protein in Escherichia coli requires a chaperone-likeprotein. EMBO J. 15:97888

57. Henrichsen J. 1983. Twitching motility. Annu. Rev. Microbiol. 37:8193

58. Hobbs M, Mattick JS. 1993. Commoncomponents in the assembly of type-4

mbriae, DNA transfer systems, lamen-tous phage and protein secretion appara-tus: a general system for the formationof surface-associated protein complexes. Mol. Microbiol. 10:23343

59. Hoelzer MA, Michod RE. 1999. DNA re-pair and the evolution of transformationin Bacillus subtilis . III. Sex with damagedDNA. Genetics. In press

60. Inamine GS, Dubnau D. 1995. ComEA,a Bacillus subtilis integral membrane pro-tein required for genetic transformation, isneeded for both DNA binding and trans-port. J. Bacteriol. 177:304551

61. Kahn ME, Smith HO. 1984. Transforma-tion in Haemophilus: a problem in mem-brane biology. J. Membr. Biol. 81:89103

62. Kaniga K, Bossio JC, Galan JE. 1994. TheSalmonella typhimurium invasion genesinvF and invG encode homologues of theAraC and PulD family of proteins. Mol. Microbiol. 13:55568



24/29

?240 DUBNAU

constitute a competence-inducible operonthat requires the product of the tfoX (sxy)gene for transcriptional activation. J. Bac-teriol. 179:481520

64. Karudapuram S, Zhao X, Barcak GJ. 1995.DNA sequence and characterization of Haemophilus influenzae dprA + , a gene re-quired for chromosomal but not plasmidDNA transformation. J. Bacteriol. 177:323540

65. Kazmierczak BI, Mielke DL, Russel M,Model P. 1994. pIV, a lamentous phageprotein that mediates phage export across

the bacterial envelope, forms a multimer. J. Mol. Biol. 238:1879866. Koster M, Bitter W, de Cock H, Allaoui A,

Cornelis GR, et al. 1997. The outer mem-brane component, YscC, of the Yop secre-tion machinery of Yersinia enterocoliticaforms a ring-shaped multimeric complex. Mol. Microbiol. 26:78997

67. Kroll JS, Wilks KE, Farrant JL, LangfordPR. 1998. Natural genetic exchange be-

tween Haemophilus and Neisseria : inter-generic transfer of chromosomal genes be-tween major human pathogens. Proc. Natl. Acad. Sci. USA 95:1238185

68. Lacks S. 1962. Molecular fate of DNA ingenetic transformation of Pneumococcus . J. Mol. Biol. 5:11931

69. LacksS. 1978. Steps in the process ofDNAbinding and entry in transformation. InTransformation , ed. SW Glover, LO But-ter. Oxford, UK: Cotswold

70. Lacks S. 1979. Uptake of circular deoxyri-bonucleic acid and mechanism of deoxyri-bonucleic acid transport in genetic trans-formationof Streptococcus pneumoniae . J. Bacteriol. 138:4049

71. Lacks S, Greenberg B. 1976. Single-strandbreakage on binding of DNA to cells inthe genetic transformation of Diplococcus pneumoniae . J. Mol. Biol. 101:25575

72. Lacks S, Greenberg B, Carlson K. 1967.Fate of donor DNA in pneumococcal trans-

Identication of a deoxyribonuclease im-plicated in genetic transformationof Diplo-coccus pneumoniae . J.Bacteriol. 123:22232

74. Lacks S, Greenberg B, Neuberger M. 1974.Role of a deoxyribonuclease in the genetictransformation of Diplococcus pneumo-niae . Proc. Natl. Acad. Sci. USA 71:23059

75. Lacks S, Neuberger M. 1975. Membranelocation of a deoxyribonuclease implicatedin the genetic transformation of Diplococ-cus pneumoniae . J. Bacteriol. 124:1321

2976. Lacks SA. 1977. Binding and entry of DNA in bacterial transformation. In Micro-bial Interactions; Receptors and Recog-nition , ed. JL Reissig. London: Chapman& Hall

77. Lacks SA, Greenberg B. 1973. Compe-tence fordeoxyribonucleic acid uptake anddeoxyribonuclease action external to cellsin the genetic transformation of Diplococ-

cus pneumoniae . J. Bacteriol. 114:15263

78. Larson TG, Goodgal SH. 1992. DonorDNA processing is blocked by a muta-tion in the com101A locus of Haemophilusinfluenzae . J. Bacteriol. 174:339294

79. Larson TG, Goodgal SH. 1991. Sequenceand transcriptional regulation of com101A ,a locus required for genetic transformationin Haemophilus inf luenzae . J. Bacteriol.173:468391

80. Lee MS, Marians KJ. 1990. DifferentialATP requirements distinguish the DNAtranslocation and DNA unwinding activ-ities of the Escherichia coli PriA protein. J. Biol. Chem. 265:1707883

81. Levine JS, Strauss N. 1965. Lag periodcharacterizing the entry of transformingdeoxyribonucleic acid into Bacillus sub-tilis . J. Bacteriol. 89:28187

82. Linderoth NA, Simon MN, Russel M.1997. The lamentous phage pIV multimer



25/29


83. Link C, Eickernjager S, Porstendorfer D,Averhoff B. 1998. Identication and char-acterization of a novel competence gene,comC , required for DNA binding and up-take in Acinetobacter sp. strain BD413. J. Bacteriol. 180:159295

84. Londo no-Vallejo JA, Dubnau D. 1993.comF , a Bacillus subtilis late competencelocus, encodes a protein similar to ATP-dependent RNA/DNA helicases. Mol. Mi-crobiol. 9:11931

85. Londo no-Vallejo JA, Dubnau D. 1994.Membrane association and role in DNA

uptake of the Bacillus subtilis PriA analogComF1. Mol. Microbiol. 13:19720586. Londo no-Vallejo JA, Dubnau D. 1994.

Mutation of the putative nucleotide bind-ing site of the Bacillus subtilis membraneprotein ComFA abolishes the uptake of DNA during transformation. J. Bacteriol.176:464245

87. Lorenz MG, Reipschlager K, WackernagelW. 1992. Plasmid transformation of natu-

rally competent Acinetobacter calcoaceti-cus in non-sterile soil extract and ground-water. Arch. Microbiol. 157:35560

88. Lorenz MG, Wackernagel W. 1994. Bacte-rial gene transfer by natural genetic trans-formation in the environment. Microbiol. Rev. 58:563602

89. LovePE, Lyle MJ, Yasbin RE. 1985. DNA-damage-inducible ( din ) loci are transcrip-tionally activated in competent Bacillussubtilis . Proc. Natl. Acad. Sci. USA 82:62015

90. Lunsford RD, Roble AG. 1997. comYA , agene similar to comGA of Bacillus sub-tilis , is essential for competence-factor-dependent DNA transformation in Strepto-coccus gordonii . J.Bacteriol. 179:312226

91. Martin PR, Hobbs M, Free PD, Jeske Y,Mattick JS. 1993.Characterization of pilQ ,a new gene required for the biogenesis of type 4 mbriae in Pseudomonas aerugi-nosa . Mol. Microbiol. 9:85768

coccus pneumoniae . Mol. Gen. Genet. 213:44448

93. Mejean V, Claverys JP. 1993. DNA pro-cessing during entry in transformation of Streptococcus pneumoniae . J. Biol. Chem.268:559499

94. Mejean V, Claverys JP. 1984. Use of acloned DNA fragment to analyze the fateof donor DNA in transformation of Strep-tococcus pneumoniae . J. Bacteriol. 158:117578

95. Michod RE, Wojciechowski MF, HoelzerMA. 1988. DNA repair and the evolution

of transformation in the bacterium Bacillussubtilis . Genetics 118:313996. Mohan S, Aghion J, Guillen N, Dubnau D.

1989. Molecular cloning and characteriza-tion of comC , a late competence gene of Bacillus subtilis . J. Bacteriol. 171:604351

97. Mongold JA. 1992. DNA repair and theevolution of transformation in Haemo- philus influenzae . Genetics 132:89398

98. Morrison DA, Guild WR. 1973. Breakageprior to entry of donor DNA in pneumo-coccus transformation. Biochim. Biophys. Acta 299:54556

99. Morrison DA, Guild WR. 1973. Structureof deoxyribonucleic acid on the cell sur-face during uptake by Pneumococcus . J. Bacteriol. 115:105562

100. Newhall WJ, Wilde CED, Sawyer WD,Haak RA. 1980. High-molecular-weightantigenic protein complex in the outermembrane of Neisseria gonorrhoeae . In- fect. Immun. 27:47582

101. Notani N,Goodgal SH. 1966.Onthenatureof recombinants formed during transfor-mation in Hemophilus influenzae . J. Gen.Physiol. 49(2):197209

102. Nunn D, Bergman S, Lory S. 1990. Prod-ucts of three accessory genes, pilB, pilC ,and pilD , are required for biogenesis of Pseudomonase aeruginosa pili. J. Bacte-riol. 172:291119



26/29

?242 DUBNAU

characterizationof natural transformationin Acinetobacter calcoaceticus . J. Gen. Microbiol. 139:295305

104. PestovaEV, Morrison DA. 1998. Isolationand characterization of three Streptococ-cus pneumoniae transformation-specicloci by use of a lacZ reporter insertionvector. J. Bacteriol. 180:270110

105. Pestova EV, Morrison DA. 1997. Strep-tococcus pneumoniae chromosomal locispecically induced during competencefor genetic transformation. Abstr. Gen. Meet. Am. Soc. Microbiol. 97th, Miami,

p. 293. Washington, DC: Am. Soc. Mi-crobiol.106. Piechowska M, Fox MS. 1971. Fate of

transformingdeoxyribonuclease in Bacil-lus subtilis . J. Bacteriol. 108:68089

107. Porstendorfer D, Drotschmann U, Aver-hoff B. 1997. A novel competence gene,comP , is essential for natural transforma-tion of Acinetobacter sp. strain BD413. Appl. Environ. Microbiol. 63:415057

108. Possot O, dEnfert C, Reyss I, Pugs-ley AP. 1992. Pullulanase secretionin Escherichia coli K-12 requires acytoplasmic protein and a putative poly-topic cytoplasmic membrane protein. Mol. Microbiol. 6:95105

109. ProvvediR, DubnauD.1998. ComEA is aDNA receptor for transformation of com-petent Bacillus subtilis . Mol. Microbiol.In press

110. Pugsley AP. 1993. The complete generalsecretory pathway in gram-negative bac-teria. Microbiol. Rev. 57:50108

111. Puyet A, Greenberg B, Lacks SA. 1990.Genetic and structural characterization of EndA. A membrane-bound nuclease re-quired for transformation of Streptococ-cus pneumoniae . J. Mol. Biol. 213:72738

112. Redeld RJ. 1993. Evolution of natu-ral transformation: testing the DNA re-pair hypothesis in Bacillus subtilis and

113. Redeld RJ. 1993. Genes for breakfast:the have-your-cake-and-eat-it-too of bac-terial transformation. J. Hered. 84:4004

114. Redeld RJ,SchragMR,Dean AM. 1997.The evolution of bacterial transformation:sex with poor relations. Genetics 146:2738

115. Reyss I, Pugsley AP. 1990. Five addi-tional genes in the pulC-O operon of thegram-negative bacterium Klebsiella oxy-toca UNF5023whicharerequired forpul-lulanase secretion. Mol. Gen. Genet. 222:17684

116. Rivas S, Bolland S, Cabezon E, Goni FM,de la Cruz F. 1997. TrwD, a protein en-coded by the IncW plasmid R388, dis-plays an ATP hydrolase activity essentialfor bacterial conjugation. J. Biol. Chem.272:2558390

117. Rosenthal AL, Lacks SA. 1977. Nucle-ase detection in SDS-polyacrylamide gelelectrophoresis. Anal. Biochem. 80:7690

118. Rosenthal AL, Lacks SD. 1980. Com-

plex structure of the membrane nucleaseof Streptococcus pneumoniae revealed bytwo-dimensional electrophoresis. J. Mol. Biol. 141:13346

119. Rudel T, Facius D, Barten R, Scheuerp-ug I, Nonnenmacher E, et al. 1995. Roleof pili and the phase-variable PilC pro-tein in natural competence for transfor-mation of Neisseria gonorrhoeae . Proc. Natl. Acad. Sci. USA 92:798690

120. Rudel T, van Putten JP, Gibbs CP, HaasR, Meyer TF. 1992. Interaction of twovariable proteins (PilE and PilC) requiredfor pilus-mediated adherence of Neisse-ria gonorrhoeae tohumanepithelial cells. Mol. Microbiol. 6:343950

121. Russel M. 1998. Macromolecular assem-bly and secretion across the bacterial cellenvelope: type II protein secretion sys-tems. J. Mol. Biol. 279:48599

122. Russel M. 1994. Phage assembly: aparadigm for bacterial virulence factor



27/29


gene product of Pseudomonas aerugi-nosa , required forpilusbiogenesis, sharesamino acid sequence identity with theN-terminus of type-4 prepilin proteins. Mol. Microbiol. 13:97385

124. Seifert HS, Ajioka RS, Paruchuri D, Hef-fron F, So M. 1990. Shuttle mutagenesisof Neisseria gonorrhoeae : pilin null mu-tations lower DNA transformation com-petence. J. Bacteriol. 172:4046

125. Shevchik VE, Robert-Baudouy J, Con-demine G. 1997. Specic interaction be-tween OutD, an Erwinia chrysanthemi

outer membrane protein of the generalsecretory pathway, and secreted proteins. EMBO J. 16:300716

126. Singh RM. 1972. Number of deoxyri-bonucleic acid uptake sites in compe-tent cells of Bacillus subtilis . J. Bacteriol.110:26672

127. Sisco KL, Smith HO. 1979. Sequence-specic DNA uptake in Haemophilustransformation. Proc. Natl. Acad. Sci.

USA 76:97276128. Smith H, Danner DB, Deich RA. 1981.

Genetic transformation. Annu. Rev. Bio-chem. 50:4168

129. Smith HO, Tomb JF, Dougherty BA,Fleischmann RD, Venter JC. 1995. Fre-quency and distribution of DNA up-take signal sequences in the Haemophilusinfluenzae Rd genome. Science 269:53840

130. Strauss N. 1965. Conguration of trans-forming deoxyribonucleic acid during en-try into Bacillus subtilis . J. Bacteriol. 89:28893

131. Strauss N. 1966. Further evidence con-cerning the conguration of transformingdeoxyribonucleic acid during entry into Bacillus subtilis . J. Bacteriol. 91:7028

132. Strauss N. 1970. Transformation of Bacil-lus subtilis using hybrid DNA moleculesconstructed by annealing resolved com-plementary strands. Genetics 66:583

single bifunctional enzyme, PilD, cata-lyzes cleavage and N-methylation of pro-teins belonging to the type IVpilin family.Proc. Natl. Acad. Sci. USA 90:24048

134. Suerbaum S, Smith JM, Bapumia K,Morelli G, Smith NH, et al. 1998. Free re-combination within Helicobacter pylori .Proc. Natl. Acad. Sci. USA 95:1261924

135. Tokai N, Fujimoto-Nishiyama A, Toyo-shima Y, Yonemura S, Tsukita S, et al.1996. Kid, a novel kinesin-like DNAbinding protein, is localized to chromo-

somes and the mitotic spindle. EMBO J.15:45767136. Tomb JF, El-Hajj H, Smith HO. 1991.

Nucleotide sequence of a cluster of genes involved in the transformation of Haemophilus influenzae RD. Gene 104:110

137. Tomb JF. 1992. A periplasmic proteindisulde oxidoreductase is required fortransformation of Haemophilus influen-

zae Rd. Proc. Natl. Acad. Sci. USA 89:1025256

138. Tnjum T, Freitag NE, Namork E,Koomey M. 1995. Identication andchar-acterization of pilG , a highly conservedpilus-assembly gene in pathogenic Neis-seria . Mol. Microbiol. 16:45164

139. Vagner V, Claverys JP, Ehrlich SD,Mejean V. 1990. Direction of DNA en-try in competent cells of Bacillus subtilis . Mol. Microbiol. 4:178588

140. van Nieuwenhoven MH, Hellingwerf KJ,Venema G, Konings WN. 1982. Role of proton motive force in genetic transfor-mation of Bacillus subtilis . J. Bacteriol.151:77176

141. Whitchurch CB,HobbsM, LivingstonSP,Krishnapillai V, Mattick JS. 1991. Char-acterization of a Pseudomonas aerugi-nosa twitching motility gene and evi-dence for a specialized protein exportsystem widespread in eubacteria. Gene



28/29

?244 DUBNAU

Michod RE. 1989. DNA repair and theevolution of transformation in Bacillussubtilis . II. Role of inducible repair.Genetics 121:41122

143. Wolfgang M, Lauer P, Park HS, Brossay

L, Hebert J, et al. 1998. PiIT muta-tions lead to simultaneous defects in com-petence for natural transformation andtwitching motility in piliated Neisseriagonorrhoeae . Mol. Microbiol. 29:32130



29/29

Annual Review of MicrobiologyVolume 53, 1999

CONTENTS

Transformation of Leukocytes by Theileria and T. annulata , Dirk Dobbelaere, Volker Heussler

1

Addiction Modules and Programmed Cell Death and Antideath inBacterial Cultures, Hanna Engelberg-Kulka, Gad Glaser

43

Wolbachia Pipientis : Microbial Manipulator of Arthropod Reproduction, R. Stouthamer, J. A. J. Breeuwer, G. D. D. Hurst

71

Aerotaxis and Other Energy-Sensing Behavior in Bacteria, Barry L.Taylor, Igor B. Zhulin, Mark S. Johnson

103

In Vivo Genetic Analysis of Bacterial Virulence, Su L. Chiang, John J. Mekalanos, David W. Holden

129

The Induction of Apoptosis by Bacterial Pathogens, Yvette Weinrauch, Arturo Zychlinsky

155

Poles Apart: Biodiversity and Biogeography of Sea Ice Bacteria, JamesT. Staley, John J. Gosink

189

DNA Uptake in Bacteria, David Dubnau 217Integrating DNA: Transposases and Retroviral Integrases, L. Haren, B.Ton-Hoang, M. Chandler

245

Transmissible Spongiform Encephalopathies in Humans, Ermias D. Belay 283

Bacterial Biocatalysts: Molecular Biology, Three-Dimensional Structures,and Biotechnological Applications of Lipases, K-E. Jaeger, B. W.

Dijkstra, M. T. Reetz315

Contributions of Genome Sequencing to Understanding the Biology of Helicobacter pylori , Zhongming Ge, Diane E. Taylor

353

Circadian Rhythms in Cyanobacteria: Adaptiveness and Mechanism, Carl Hirschie Johnson, Susan S. Golden

389

Constructing Polyketides: From Collie to Combinatorial Biosynthesis,

Ronald Bentley, J. W. Bennett 411Giant Viruses Infecting Algae, James L. Van Etten, Russel H. Meints 447Mechanisms for Redox Control of Gene Expression, Carl E. Bauer,Sylvie Elsen, Terry H. Bird

495

Intercellular Signaling During Fruiting-Body Development of Myxococcus xanthus , Lawrence J. Shimkets

525

Clostridial Toxins as Therapeutic Agents: Benefits of Nature's Most ToxicProteins, Eric A. Johnson

551

Viruses and Apoptosis, Anne Roulston, Richard C. Marcellus, Philip E. Branton

577

The Cytoskeleton of Trypanosomatid Parasites, Keith Gull 629

Documents

Dubnau 1999 ARM (Review - DNA Uptake in Bacteria)